This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
precision = Precision() | |
precision.update_state(y_train, y_train_pred) | |
precision.result().numpy() |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from sklearn.metrics import confusion_matrix | |
import seaborn as sns | |
# notice the threshold | |
def plot_cm(labels: numpy.ndarray, predictions: numpy.ndarray, p: float=0.5) -> (): | |
cm = confusion_matrix(labels, predictions > p) | |
# you can normalize the confusion matrix | |
plt.figure(figsize=(5,5)) | |
sns.heatmap(cm, annot=True, fmt="d") |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Evaluate the model on the test data using `evaluate` | |
print("Evaluate on test data") | |
score_test = model.evaluate(test_ds.batch(batch_size)) | |
for name, value in zip(model.metrics_names, score_test): | |
print(name, ': ', value) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
def plot_metrics(history: History) -> (): | |
metrics = ['loss', 'precision', 'recall', 'auc', 'tp', 'sensitivity'] | |
for n, metric in enumerate(metrics): | |
name = metric.replace("_"," ").capitalize() | |
plt.subplot(3, 2, n+1) # adjust according to metrics | |
plt.plot(history.epoch, history.history[metric], color=colors[0], label='Train') | |
plt.plot(history.epoch, history.history['val_'+metric], | |
color=colors[0], linestyle="--", label='Val') | |
plt.xlabel('Epoch') | |
plt.ylabel(name) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import matplotlib.pyplot as plt | |
from matplotlib import rcParams | |
rcParams['figure.figsize'] = (12, 10) | |
colors = plt.rcParams['axes.prop_cycle'].by_key()['color'] | |
def plot_log_loss(history: History, title_label: str, n: int) -> (): | |
# Use a log scale to show the wide range of values. | |
plt.semilogy(history.epoch, history.history['loss'], | |
color=colors[n], label='Train '+title_label) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
batch_size = 64 | |
""" | |
Training the model for 60 epochs using our dataset. | |
The batch size (64) is the same for the validation data. | |
Only 1 callback was used, but could be more like TensorBoard, ModelCheckpoint, etc. | |
""" | |
history = model.fit(train_ds.batch(batch_size=batch_size), | |
epochs=60, | |
validation_data=validation_ds.batch(batch_size=batch_size), |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from tensorflow.keras.callbacks import EarlyStopping | |
""" | |
This callback will stop the training when there is no improvement in the validation accuracy across epochs | |
""" | |
early_callback = EarlyStopping(monitor='val_auc', | |
verbose=1, | |
patience=10, | |
mode='max', | |
restore_best_weights=True) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from tensorflow.keras.losses import BinaryCrossentropy | |
from tensorflow.keras.optimizers import Adam | |
from tensorflow.keras.metrics import TruePositives, FalsePositives, TrueNegatives, FalseNegatives, BinaryAccuracy, Precision, Recall, AUC | |
from tensorflow.keras.metrics import SpecificityAtSensitivity | |
""" | |
Definition of metrics commonly used on medical imaging classification, segmentation, and localization problems. | |
The metrics will appear on each iteration of the training process to monitor the progress of our design. | |
""" | |
METRICS = [ |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from tensorflow.keras.utils import plot_model | |
plot_model(model, 'my-CNNmodel.png', show_shapes=True) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from tensorflow.keras import Model | |
from tensorflow.keras.layers import Input, Convolution2D, MaxPool2D, BatchNormalization, Flatten, Dropout, Dense | |
from tensorflow.keras.regularizers import l2 | |
from tensorflow.keras.activations import relu, sigmoid | |
from tensorflow.keras.initializers import GlorotNormal | |
# this configuration uses backend.set_image_data_format('channels_first') | |
""" | |
This design creates the same network than before but using the layer by layer configuration. |